AI Proves That It Can Be Creative With Whatever Text Prompts You Throw At It
Images via OpenAI
Throw lemons, along with a few other hints, at artificial intelligence, and it will make lemonade.
Tackling the question about whether AI can replicate human creativity, Microsoft-backed research nonprofit OpenAI has developed a neural network called DALL·E that remarkably creates images out of text prompts. The answer seems to be that it’s getting there.
DALL·E, who is named after both Pixar’s WALL-E robot and surrealist art legend Salvador Dalí, is able to comprehend text captions in natural language and then interpret them from scratch in a wide spectrum of concepts. These include photography in fashion and interior design settings, whimsical emojis, and the mixing of real-world and imaginary ideas.
One incredible example is a series of drawings churned out of the prompt, “An illustration of a baby daikon radish in a tutu walking a dog.”
Images via OpenAI
Also, check out these versions of a “giraffe turtle chimera”:
Images via OpenAI
DALL·E’s creators say that, with the need to understand multiple objects and attributes simultaneously, the neural network faces the added challenge of understanding phrasing so that it doesn’t mix any of the attributes up.
“For example, consider the phrase ‘a hedgehog wearing a red hat, yellow gloves, blue shirt, and green pants.’ To correctly interpret this sentence, DALL·E must not only correctly compose each piece of apparel with the animal, but also form the associations (hat, red), (gloves, yellow), (shirt, blue), and (pants, green) without mixing them up,” the OpenAI team shares.
Given the complexity of language, DALL·E still has its limitations. It is able to comprehend strings of input listing “a small number of objects,” but when its developers introduce more items, “DALL·E is prone to confusing the associations between the objects and their colors, and the success rate decreases sharply.”
Noting the flimsy standing of AI in ethics, OpenAI also hopes to study how models like DALL·E can bring improvements to the societal and economic scapes.
To view more images created by the neural network, please head here.
Text prompt: “A living room with two white armchairs and a painting of the Colosseum. The painting is mounted above a modern fireplace.”
Images via OpenAI
Text prompt: “A snail made of harp. A snail with the texture of a harp.”
Images via OpenAI
Text prompt: “A professional high quality emoji of a lovestruck cup of boba.”
Images via OpenAI
Text prompt: “The exact same cat on the top as a sketch on the bottom.”
Images via OpenAI
Text prompt: “A painting… of a capybara sitting in a field at sunrise.”
Images via OpenAI
[via Forbes, images via OpenAI]
Also check out these recent news
